Interrupted time series studies: investigation of methods for analysis, synthesis and interpretation

Interrupted time series (ITS) studies are frequently used to evaluate the effects of policy and health systems interventions. These designs can be employed in circumstances where it is inappropriate or impossible to use a randomised trial. Evidence from ITS studies is commonly incorporated in systematic reviews, which, in turn, underpin public health and clinical practice guidelines, and inform policy decisions. Despite the increasing popularity of the design, there has been surprisingly little evaluation of the statistical methods used to analyse such designs, or how the results from multiple ITS studies should be quantitatively synthesized (a method known as meta-analysis). Nor has there been development or evaluation of formats to present results from individual ITS studies, or the synthesis of ITS results, to facilitate interpretation by policy makers. The aims of this project are: 1) To provide evidence-informed guidance and tools for researchers on the design, analysis and meta-analysis of ITS studies. 2) Support interpretation of analyses and meta-analyses of ITS studies in policy decision-making.